Co-Creation in Contextual Competences for Sustainability: Teaching for Sustainability, Student Interaction and Extracurricular Engagement

Abstract

Due to the need to achieve the sustainable development of society, the importance of sustainable development competences has reached unprecedented heights. This paper constructed an SEM model and considered the factors of contextual competence for sustainability from teaching, student interaction and extracurricular engagement, by conducting a questionnaire survey among 786 Chinese engineering undergraduates. The results indicated a significant and direct positive influence of teaching on contextual competences for sustainability. In addition, there is an indirect effect on contextual competence for sustainable development through student interaction and extracurricular engagement. An examination of the interlocking mediating effects of student inter-action and extracurricular engagement revealed that student interaction contributes to the development of students’ contextual competence for sustainable development through its positive influence on extracurricular engagement. This paper provides a reference for the cultivation of contextual competences for sustainability from both theoretical and practical perspectives.

1. Introduction

The 17 sustainable development goals (SDGs) in the Education for Sustainable Development Goals—Learning Goals describe the main challenges to human development [1]. Sustainable development education aims to train the competences of individuals to cope with the current and future impacts of society [2]. SDGs and sustainable development education are combined by reforming the teaching content and method, so that the individual can change behavior and promote the social reform [3]. Sustainable development education plays a catalytic role in the realization of the SDGs. This study focuses on how sustainable development education in higher education can help engineering students improve their competences to address complex societal issues.

The National Academy of Engineering anticipates imminent changes in the work environment, and future engineers should have the ability to be aware of complex contextual information about society [4]. Contextual competences are among the key competences that engineering students should possess. Based on the previous studies, it refers to engineering students’ ability to anticipate and recognize social and other factors’ impacts and constraints when they solve engineering problems in terms of sustainable development [5]. It is a necessary learning achievement for engineering students [6]. The Accreditation Board for Engineering and Technology (ABET) states in its Student Outcomes that students must consider the impact of engineering solutions in the environment to make informed judgments [7]. The ABET criteria ensure that engineering graduates develop nontechnical skills while deepening their understanding of the context [8]. This helps to cultivate responsible engineering professionals of the future.

Contextual competences require engineering students to consider the constraints of the context and the feasibility of the solution. For example, what are the local impacts of the solutions? What factors will limit the implementation of the program [9]? In this way, a healthy engineering environment mutually beneficial for the sustainable development of society can be created. At present, an increasing number of engineering educators are paying attention to the importance of contextual competences, but there are still some problems. For example, engineering students lack confidence in self-assessment of their contextual competences [10], senior engineering students lack a broader and deeper understanding of the background that needs to be considered for design issues [11], or almost half of employers think engineering students are ill-prepared in this competence [12,13]. It is critical to strengthen the cultivation of contextual competences for the sustainability of engineering students.

The cultivation of contextual competencies for sustainability in higher education has been noticed and studied by several scholars. However, the existing research has some shortcomings that need to be improved. In terms of research objects, although the scope was relatively broad, involving middle school students [14,15] and college students [16,17], only a few studies have focused on engineering undergraduates in higher education. As the reserve force of engineers, undergraduate engineering students are one of the achievements of sustainable development education. Meanwhile, engineering undergraduates have a complete and mature training system for undergraduate engineering students. Therefore, the measurement of their contextual competences for sustainability is not only a test of the achievement of sustainable development education, but also ensures that engineering students play an effective role in solving future contextual engineering problems. In terms of research content, the existing research mostly has explored the influencing factors of contextual competences for sustainability from individual factors, such as curriculum experience [9], or focused on the relationship between contextual competences and the learning environment [18]. These studies have paid limited attention to the development pathway of students’ contextual competence. There is a lack of relevant systematic empirical research in this area. In addition, fewer scholars have taken into account the mediating effects of extracurricular engagement or student interaction when exploring the effects of teaching and learning on contextual competencies for sustainable development. This study is improved from the perspective of research objects and variables.

This paper constructs a model of contextual competences for sustainability and primarily addresses these four key issues: (1) to study how teaching influences sustainability competences; (2) to explore the mediating influence of extracurricular engagement between teaching and contextual competences for sustainability; (3) to investigate how student interaction mediates the relationship between teaching and sustainability competences; (4) to explore the influence of student interaction on extracurricular engagement. Based on the results of a survey involving 786 engineering undergraduates, this study explores the current state and training path for sustainability competences in sustainable development education. It offers policy recommendations to enhance sustainability competences in higher engineering education and provides theoretical countermeasures for social development and progress from the perspective of engineering education.

2. Research Hypothesis

2.1. Influence of Teaching

In this study, “teaching” is defined as the teaching methods that teachers use in the classroom. The existing literature has confirmed that teachers can better promote the sustainable development of students by using teaching methods. Lambrechts et al. [19] classified teaching methods that develop the sustainable development competences of students into three categories: interactive and participatory methods, such as group discussion and role playing; action-oriented approaches, such as internships and outdoor learning; and research methods, such as studying the literature and cases. The role of interactive and participatory teaching methods and research methods has been further confirmed. Wang et al. [20] found that case study and interdisciplinary team teaching have a positive impact on students’ thinking and perceptions of sustainable development when studying the impact of curriculum pedagogy on sustainability. Some researchers, such as Martínez Casanovas et al. [21] and Ozis et al. [22], proposed that participatory methods include problem-based and experiential learning, e.g., by adopting active teaching methods in class, students engage in the learning process and avoid passive acceptance of knowledge, which is conducive to improving students’ abilities to solve problems and so on [23]. Lozano et al. [24] explored the relationship between teaching and sustainable development competences. They found that case studies, problem-based learning and participatory research have strong applicability to the development of sustainable development competences. Khairiyah Mohd et al. [23] used a problem-based cooperative learning teaching method in an engineering course and found that engineering students had significantly improved their problem-solving and teamwork abilities related to sustainable development. This proves the effect of various teaching methods on cultivating sustainable development competences.

Given the existing research on teaching, the following hypothesis is proposed in this study:

H1: 

Teachinghas a significant positive and direct influence on engineering students’ contextual competences for sustainability.

2.2. The Mediating Roles of Extracurricular Engagement and Student Interaction

“Extracurricular” activities refer to learning activities outside the formal curriculum [25,26]. Extracurricular engagement can be understood as the time and energy invested by students in activities besides the formal curriculum.

Research by Pike et al. [27] found significant direct relationships between undergraduate education spending and learning outcomes, as well as significant indirect effects on broader outcomes through various forms of student engagement. In the classroom, students’ perception of supportive teaching adopted by teachers can have a positive impact on students’ extracurricular participation activities by influencing their autonomous motivation [28]. In terms of extracurricular engagement, what students do outside the classroom shapes their attitudes and learning outcomes in subtle and complex ways. Engineering has high requirements of students in terms of difficulty, depth and breadth, which require that students not only to absorb knowledge in class, but also to make great efforts after class [29]. Extracurricular learning is therefore considered to have a positive impact by encouraging deep learning, improving students’ problem-solving skills, analytical and critical skills, and confidence, indicating that there is a direct relationship between students’ extracurricular learning and individual progress in higher education. Some researchers believed that activities that students engage in outside the classroom provide a real context for sustainable development, which is conducive to the cultivation of competences [30].

Based on the existing research on extracurricular engagement, the second research hypothesis of this paper is as follows:

H2: 

Extracurricular engagement mediates the influence of teaching on contextual competences for sustainability.

Student interaction promotes the construction of knowledge [31], and has an impact on the development of students. Studies have shown that students share their ideas with peers during the writing process and receive constructive comments from them. This helps students reduce writing anxiety and improve their writing skills [32]. Leslie’s [33] research showed that group projects help students gain insight into what they are learning. Students support each other and exchange ideas in cooperation. Course quality is influenced by cooperation and reciprocity among learners. When learners engage with peers, they experience a deeper level of learning and satisfaction [34]. It can be seen that the interaction between learners plays an important role in student development.

A teaching method with cooperation and communication as the theme provided opportunities and an environment for interaction between students and had an impact on their development [21]. The research showed that adopting cooperative and interactive teaching methods can produce significant positive effects on students’ writing skills [35], academic achievement [36] and engagement in class [21]. Teachers provided a platform for student interaction and guidance for peer feedback. Gradually, students realized that it is an effective learning strategy that can be used to motivate each other and help achieve learning goals [37].

Based on the existing research on student interaction, the third research hypothesis of this paper is as follows:

H3: 

Student Interaction mediates the influence of teaching on contextual competences for sustainability.

2.3. The Chain Mediating Roles of Student Interaction and Extracurricular Engagement

Many factors influence student engagement, such as campus environment, institutions and peer effect [38]. As an important subject of students’ contact, peers’ words and deeds have an impact on students’ behavior. Research showed that perceived support from peers has a positive predictive effect on students’ prosocial behavioral motivation. The social behavior of students also affected peers’ behaviors, but the quality of friendship determines the degree of impact [39].

It is known from previous studies that student interaction has a significant effect on the level of student engagement. Bucea-Manea-Țoniş et al. [40] found that communication among students also increases students’ engagement in class. In addition, student interaction was also an important driver of students’ engagement in game-based activities [41]. Interaction between students also promotes students’ interest in learning. When students are motivated by influential peers, they tend to develop positive attitudes towards activities, which increases their engagement [42]. Some teaching methods provide more opportunities for student communication [23].Li et al. [43] found that learning methods including student–student interaction have an impact on students’ social and behavioral engagement.

Based on the existing research on student interaction and extracurricular engagement, the following hypothesis is proposed in this paper:

H4: 

Student interaction and extracurricular engagement have chain mediating roles in the mediates influence of teaching on contextual competences for sustainability.

The current theoretical framework is shown in Figure 1.

3. Research Methods

This study was divided into three steps: Firstly, descriptive statistics were performed to describe the frequency tables of the categorical variables and the normality check of the continuous variables. Secondly, confirmatory factor analyses were conducted, which included unstandardized factor loadings, standard errors, significance estimates and standardized factor loadings. Composite reliability was computed, and convergent validity was computed for the AVE and discriminant validity of the Fornell and Larcker criteria. Lastly, the structural model was used to calculate the model fit, the significance of the regression coefficients, and the mediating effects.

3.1. Survey Object and Questionnaires Collection

In this study, we selected a sample of senior undergraduates majoring in engineering in Beijing, whose majors involve materials engineering, mechanical engineering and so on. In order to ensure the effectiveness of this questionnaire survey and the validity of the survey tools, before the formal distribution of questionnaires, we made a pilot survey and distributed 100 questionnaires to fourth-year engineering undergraduates in engineering colleges. After eliminating ineffective questionnaires such as too short answering time and regular answering, there were 83 valid ones. We conducted an item analysis on the collected questionnaires and revised the questions according to the analysis results. During the formal investigation, we randomly distributed the revised questionnaires online to survey objects. Finally, 830 questionnaires were collected. After exclusion of invalid questionnaires, there were 786 valid ones. In the process of distributing questionnaires, informed consent from the respondents was obtained. All surveys were conducted on a voluntary basis. The confidentiality and anonymity of personal information were guaranteed.

3.2. Survey Tools and Statistical Results of Variables

This study began with a thorough literature study, collecting and organizing relevant mature scales from home and abroad. Based on this, adapted to the aims and objectives of this study, the survey tools designed include the following two aspects: The first is the demographic variables of the survey subjects, including gender, place of birth, parents’ education level and annual household income.

The detailed analysis results of demographic variables of research objects are shown in

Table 1. Analysis results of demographic variables of research objects.

Variable	Value Label	Value	Frequency	Valid Percent	Cum Percent
Gender	Male	1	382	48.60	48.60
	Female	2	404	51.40	100.00
		Total	786	100.0
Birthplace	Cities and towns	1	511	65.01	65.01
	Rural	2	275	34.99	100.00
		Total	786	100.0
Education level of parents	Junior high or under	1	108	13.74	13.74
	Senior high school	2	201	25.57	39.31
	Higher vocational school	3	134	17.05	56.36
	College	4	335	42.62	98.98
	Graduate school	5	8	1.02	100.00
		Total	786	100.0
Family annual income	Under 30 K	1	25	3.18	3.18
	30–80 K	2	162	20.61	23.79
	80–150 K	3	259	32.95	56.74
	150–300 K 300 K or above	4 5	247 93	31.42 11.83	88.17 100.00
		Total	786	100.0

Note: Cum Percent: cumulative percent.

.

The second aspect is the measurement of sustainability and its influencing factors among engineering students. It was measured using a 7-point Likert scale from 1 to 7 where “very inconsistent” is 1, “average” is 4 and “very consistent” is 7. The question content and reference of indicators are shown in

Table 2. Question content and reference of indicators.

Indicators	Content	Reference
Contextual competence	Ability to anticipate and recognize social and other factors’ impacts and constraints when they solve engineering problems in terms of sustainability.	ABET [7], WA [44]
Teaching for sustainability	Inquiry teaching, case teaching, project teaching;Group cooperation, hands-on practice, use of simulation software.	Lattuca, Terenzini, Knight and Ro [45]
		Zhuang, Cheung, Lau and Tang [46]
Extracurricular engagement	Students’ participation in innovation and entrepreneurship training, social services, placement, etc.	Zongyu Hu [47]
		Yan Tan [48]
Student interaction	Cooperate to complete coursework, help students understand knowledge, consult classmates, and share learning experiences	Zilvinskis [49]

. The first is a measure of the survey respondents’ contextual competences for sustainable development, and the scale consists of five questions.

This study integrates the requirements of “the Washington Accord” [44] and the Accreditation Board for Engineering and Technology [7] in terms of sustainability and context, such as, “I have a clear understanding of the industrial policies and laws and regulations of my major to ensure the legal compliance of engineering activities”.

Another part of the second aspect of the survey tool is the measurement of factors influencing the contextual competences for sustainability. The scale of factors affecting the contextual competences for sustainability includes three aspects: teaching, extracurricular engagement and student interaction. In terms of teaching, Lattuca and Zhuang’s [45,46] teaching scale and group learning scale are adopted in this study, and revised according to the actual situation by referring to other Chinese and English literature sources. Finally, the teaching scale in this study involves five questions, mainly asking students what teaching methods teachers use in class, including inquiry teaching, case teaching, group cooperation, etc.

To measure the degree of extracurricular engagement of students, this study clarified the indicators of students’ extracurricular participation based on a literature analysis [47,48] and revised them according to the actual situation. Finally, extracurricular engagement involves four questions, which ask students about extracurricular engagement in competitions, voluntary service, etc. For example, “I took the initiative to engage in volunteer activities related to engineering.

The survey was based on the National Survey of Student Engagement (NSSE) [49] and other references related to student interaction, which were revised according to the particularity of the research content. Finally, the student interaction scale involved four questions, including cooperation, assistance in understanding, advice on difficulties, and sharing of the learning experience. For example, “I will actively cooperate with other students to complete coursework or projects.”

This study used SPSS 25.0 and AMOS 24.0 statistical software to analyze the descriptive statistics and model analysis of the variables. AMOS 24.0 software was used to test the measurement model, determine the fit degree and reliability, then test the variable path. The bootstrap method was adopted to examine the mediating effect of extracurricular engagement and student interaction.

4. Results

4.1. Descriptive Statistics

The descriptive results are presented in

Table 3. Descriptive statistical.

Variable	N	Mean	Std Dev	Kurtosis	Skewness	Minimum	Maximum
TEAC01	786	5.53	1.27	−0.77	−0.42	2	7
TEAC02	786	5.54	1.26	−0.66	−0.46	2	7
TEAC03	786	5.54	1.26	−0.77	−0.44	2	7
TEAC04	786	5.52	1.28	−1.11	−0.29	2	7
TEAC05	786	5.54	1.26	−0.82	−0.36	1	7
STUD01	786	5.33	1.01	−0.44	−0.14	2	7
STUD02	786	5.53	1.10	−0.54	−0.31	2	7
STUD03	786	5.67	1.18	−0.87	−0.39	2	7
STUD04	786	5.59	1.20	−1.01	−0.29	2	7
EXTR01	786	5.79	1.25	−0.81	−0.58	2	7
EXTR02	786	5.70	1.27	−0.93	−0.50	2	7
EXTR03	786	5.66	1.31	−1.12	−0.42	2	7
EXTR04	786	5.59	1.31	−0.88	−0.46	2	7
SUST01	786	5.74	1.23	−0.88	−0.49	2	7
SUST02	786	5.82	1.23	−0.73	−0.65	2	7
SUST03	786	5.86	1.25	−0.47	−0.78	2	7
SUST04	786	5.61	1.33	−1.05	−0.47	2	7
SUST05	786	6.31	0.97	0.59	−1.21	2	7

Note: TEAC: teaching; EXTR: extracurricular engagement; STUD: student interaction; SUST: contextual competences for sustainability.

. The range of the mean is from 5.33 to 6.31, the range of standard deviations is from 0.97 to 1.33, the range of skewness values is from −1.21 to −0.14 and the range of kurtosis values are from −1.12 to 0.59. Most of observed variables meet the suggestions by Kline [50], that the absolute skewness value must be less than two and absolute kurtosis value less than seven; most of the variables conform to the normal distribution.

4.2. Measurement Model Analysis

4.2.1. Convergent Validity

This research is divided into two parts: one is the measurement model and the other is the structural model according to the suggestion of Anderson and Gerbing [51], respectively. The purpose of the measurement model is to test the reliability and validity of the research constructs. Confirmatory factor analysis (CFA) is the most suitable statistical technique, and the purpose of the structural model is testing the sign and strength of regression weights included the significance of regression coefficients. This study adopted the maximum likelihood estimation (MLE) as the estimator, which is also the estimator embedded in SEM. The analysis that the measurement model provides includes the strength of unstandardized factor loadings, significance test, standardized factor loadings, reliability (internal consistency), convergent validity and discriminant validity.

The analysis results of the measurement model are shown in the table below. The third column in

Table 4. Results for the measurement model.

Construct	Item	Significance of Estimated Parameters				Item Reliability		Construct Reliability	Convergence Validity
		Unstd.	S.E.	Unstd./S.E.	p-Value	Std.	SMC	CR	AVE
TEAC	TEAC01	1.000				0.675	0.456	0.814	0.467
	TEAC02	1.000	0.064	15.729	0.000	0.681	0.463
	TEAC03	1.062	0.065	16.434	0.000	0.722	0.521
	TEAC04	1.010	0.065	15.642	0.000	0.676	0.457
	TEAC05	0.976	0.063	15.376	0.000	0.661	0.437
EXTR	EXTR01	1.000				0.616	0.379	0.770	0.457
	EXTR02	1.114	0.081	13.794	0.000	0.674	0.454
	EXTR03	1.173	0.084	13.943	0.000	0.687	0.471
	EXTR04	1.233	0.086	14.308	0.000	0.723	0.523
STUD	STUD01	1.000				0.673	0.453	0.767	0.451
	STUD02	1.108	0.076	14.534	0.000	0.684	0.468
	STUD03	1.187	0.082	14.468	0.000	0.679	0.461
	STUD04	1.156	0.082	14.088	0.000	0.651	0.423
SUST	SUST01	1.000				0.763	0.582	0.878	0.590
	SUST02	0.987	0.047	20.944	0.000	0.751	0.563
	SUST03	1.000	0.048	20.899	0.000	0.749	0.561
	SUST04	1.059	0.051	20.756	0.000	0.744	0.554
	SUST05	0.855	0.037	23.241	0.000	0.829	0.686

Note 1: Unstd.: unstandardized factor loadings; Std: standardized factor loadings; SMC: square multiple correlations; CR: composite reliability; AVE: average variance extracted. Note 2: TEAC: teaching; EXTR: extracurricular engagement; STUD: student interaction; SUST: contextual competence for sustainability.

is unstandardized factor loadings, the fourth column is standard errors, and the fifth column is unstandardized factor loadings divided by standard errors, also known as the z- or t-value. The sixth column is the p-value of the significance test, the seventh column is standardized factor loadings, the eighth column is square multiple correlations (the square of factor loading), the ninth column is composite reliability and the last column is average variance extracted (AVE). Fornell and Larcker [52] suggested that the convergent validity evaluation of a construct includes, firstly, the reliability of all items, or alternatively, square of factor loading (square multiple correlations—SMC). Secondly, the composite reliability of each dimension, and finally, the average variance extracted (AVE), are also included. Composite reliability is similar to Cronbach’s alpha (internal consistency)—both can represent reliability, but normally speaking, composite reliability (CR) is similar to Cronbach’s alpha (internal consistency); both can represent reliability, but normally speaking, CR in SEM is more commonly used.

From the following table, standardized factor loadings of items are between 0.616 and 0.829, all values are within a reasonable limit. The constructs’ composite reliability’s smallest value is 0.767, exceeding the 0.7 recommended by Nunnally and Bernstein [53], indicating they have good internal consistency.

4.2.2. Discriminant Validity

Following Fornell and Larcker’s [52] criteria, the diagonal is the square root of the average variance extracted (√AVE), and the lower triangle is the Pearson correlation between the construct. If √AVE is larger than other correlations, then discriminant validity exists [54].

From

Table 5. Discriminant validity of the measurement model.

	AVE	TEAC	EXTR	STUD	SUST
TEAC	0.467	0.683
EXTR	0.457	0.215	0.676
STUD	0.451	0.224	0.215	0.672
SUST	0.590	0.383	0.428	0.317	0.768

Note 1: The bold numbers on the diagonal are square roots of the AVE; off-diagonal elements are the correlation of constructs. Note 2: TEAC: teaching; EXTR: extracurricular engagement; STUD: student interaction; SUST: contextual competences for sustainability.

, all constructs’ √AVE are larger than the correlations of related constructs. According to the results of Table 4, this provides sufficient evidence for the existence of discriminant validity.

4.3. Structural Model Analysis

4.3.1. Model Goodness-of-Fit Test and Path Coefficients Results

Maximum likelihood estimation (MLE) is the most common method and default in structural equation modeling software. Our study had 786 respondents and latent variables, which is suited for using structural equation modeling to test the research hypothesis and model fit. In addition to providing parameter significance estimates, SEM also provides model fit to understand how well the proposed model fits the data. Kline [50], Schumacker and Lomax [55] explained the meaning of numerous fit indices to use in a structural equation model. Jackson, Gillaspy Jr. and Purc-Stephenson [56] reviewed one hundred and ninety-four structural equation model (SEM) studies in journals of the American Psychological Association (APA) from 1998 to 2006. From past experience, it was found that there are several fit indices commonly reported in SEM. They are χ2, degree of freedom (df), χ2/df, RMSEA, SRMR, CFI, TLI (NNFI), GFI, AGFI, etc.

Good models have an RMSEA of 0.05 or less. Models whose RMSEA is 0.10 or more have poor fit. This measure is the standardized difference between the observed covariance and predicted covariance. A value of zero indicates perfect fit. This measure tends to be smaller as sample size increases and as the number of parameters in the model increases. A value less than 0.08 is considered a good fit.

Table 6. Model fit.

Model Fit	Criteria	Model Fit of Research Model
MLχ2	The smaller the better	229.014
DF	The larger the better	129.000
Normed Chi-sqr (χ2/DF)	1 < χ2/DF < 3	1.775
RMSEA	<0.08	0.031
SRMR	<0.08	0.031
TLI (NNFI)	>0.9	0.976
CFI	>0.9	0.979
GFI	>0.9	0.968
AGFI	>0.9	0.958

presents the model fits in our model as well as the thresholds of recommendation. There is no golden rule for χ2 and df, since they increase along with the complexity of the model and the number of measured items. Meanwhile, the chi-square increases when the sample size increases. Joreskog [57] proposed that χ2 be adjusted by the degrees of freedom to assess model fit. Using χ2/df as a fit index, ideally, χ2/df should be between one and three, representing a good model fit.

The fit index is presented in Table 6; all the fit indices meet the criteria suggested by Schumacker and Lomax [58]. Therefore, the model has a good fit.

The path coefficients and statistical significance are shown in

Table 7. Regression coefficient.

DV	IV	Unstd.	S.E.	Unstd./S.E.	p-Value	Std.	R2
STUD	TEAC	0.178	0.037	4.834	0.000	0.224	0.050
EXTR	TEAC	0.157	0.042	3.738	0.000	0.176	0.075
	STUD	0.198	0.055	3.622	0.000	0.175
SUST	TEAC	0.297	0.046	6.466	0.000	0.271	0.304
	EXTR	0.403	0.055	7.355	0.000	0.330
	STUD	0.257	0.058	4.433	0.000	0.186

Note: TEAC: teaching; EXTR: extracurricular engagement; STUD: student interaction; SUST: contextual competences for sustainability.

. TEAC→STUD (b = 0.178, s.e. = 0.037, p < 0.001) is significant. TEAC→EXTR (b = 0.157, s.e. = 0.042, p < 0.001) and STUD→EXTR (b = 0.198, s.e. = 0.055, p < 0.001) are also significant. TEAC→SUST (b = 0.297, s.e. = 0.046, p < 0.001), EXTR→SUST (b = 0.403, s.e. = 0.055, p < 0.001) and STUD→SUST (b = 0.257, s.e. = 0.058, p < 0.001) are all significant.

All the research hypotheses are supported. The 7.5% variance in EXTR can be explained with TEAC and STUD. The 5% variance in STUD can be explained with TEAC. The 30.4% variance of SUST can be explained with TEAC, EXTR and STUD.

4.3.2. Analysis of Mediation Effects

The research statistical model is shown in Figure 2. From the analysis of the indirect effects table shown below, the total effect of TEAC→EXTR (b = 0.192, s.e. = 0.048, p < 0.001) and the confidence interval (CI) of the bias-corrected percentile did not include 0 (CI of TEAC→EXTR = [0.102 0.293]). The total effect was supported by evidence. The total indirect effect TEAC→STUD→EXTR (b = 0.035, s.e. = 0.015, p = 0.010 < 0.05) and the bias-corrected confidence interval (CI) did not include 0 (CI of TEAC→STUD→EXTR = [0.013 0.074]). The total indirect effect is supported.

The total effect TEAC→SUST (b = 0.42, s.e. = 0.049, p < 0.001) and the bias-corrected percentile confidence interval (CI) did not include 0 (CI of TEAC→SUST = [0.326 0.520]). The total effect is supported. The specific indirect effect TEAC→EXTR→SUST (b = 0.063, s.e. = 0.002, p < 0.001) and the bias-corrected CI did not include 0 (CI of TEAC→EXTR→SUST = [0.029 0.108]). Thus, the hypothesis of specific indirect effect is supported. The specific indirect effect TEAC→STUD→SUST (b = 0.046, s.e. = 0.016, p = 0.001 < 0.05) and the bias-corrected CI did not include 0 (CI of TEAC→STUD→SUST = [0.020 0.083]). Thus, the hypothesis of specific indirect effect is supported. The indirect effect of TEAC→STUD→EXTR→SUST (b = 0.014, s.e. = 0.006, p = 0.010 < 0.05) and the bias-corrected percentile CI did not include 0 (CI of TEAC→STUD→EXTR→SUST = [0.005 0.032]). Thus, the hypothesis of specific indirect effect is supported.

The total effect STUD→SUST (b=0.337, s.e. = 0.063, p < 0.001) and the bias-corrected CI did not include 0 (CI of STUD→SUST = [0.206 0.461]). The existence of total effect is supported. The total indirect effect STUD→EXTR→SUST (b = 0.080, s.e. = 0.025, p = 0.001 < 0.05) and the bias-corrected percentile CI did not include 0 (CI of STUD→EXTR→SUST = [0.036 0.131]). The total indirect effect is supported. The analysis of indirect effects is shown in

Table 8. The analysis of indirect effects.

Effect	PointEstimate	Product of Coefficients			Bootstrap 1000 Times
					Bias-Corrected 95%
		S.E.	Z-Value	p-Value	Lower Bound	Upper Bound
Total effect
TEAC→EXTR	0.192	0.048	4.000	0.002	0.102	0.293
Total indirect effect
TEAC→STUD→EXTR	0.035	0.015	2.333	0.001	0.013	0.074
Direct effect
TEAC→EXTR	0.157	0.046	3.413	0.002	0.073	0.255
Total effect
TEAC→SUST	0.420	0.049	8.571	0.002	0.326	0.520
Total indirect effect
TEAC→SUST	0.123	0.028	4.393	0.002	0.074	0.185
Specific indirect effect
TEAC→EXTR→SUST	0.063	0.020	3.150	0.002	0.029	0.108
TEAC→STUD→SUST	0.046	0.016	2.875	0.001	0.020	0.083
TEAC→STUD→EXTR→SUST	0.014	0.006	2.333	0.001	0.005	0.032
Direct effect
TEAC→SUST	0.297	0.047	6.319	0.002	0.204	0.388
Total effect
STUD→SUST	0.337	0.063	5.349	0.002	0.206	0.461
Total indirect effect
STUD→EXTR→SUST	0.080	0.025	3.200	0.002	0.036	0.131
Direct effect
STUD→SUST	0.257	0.062	4.145	0.002	0.129	0.378

Note: TEAC: teaching; EXTR: extracurricular engagement; STUD: student interaction; SUST: contextual competences for sustainability.

.

4.3.3. Gender for Multigroup Comparison

Gender is a binary variable; it is easy, using Amos, to test regression weight difference between genders by using multigroup analysis. The prerequisite of comparing the difference between regression weights is that the measurement weights (factor loadings) should be constrained equal, indicating that men and women have the same view on the items [56]. If the measurement weights are equal, we consider the regression weights as equal between genders. If χ² is significant, it means gender has different influence in the model, and vice versa, so there is no different influence in the model.

The results of the multigroup comparison are shown in

Table 9. Comparison regressing coefficients between gender.

Model	NPAR	χ2	DF	ΔDF	Δχ2	P
Unconstrained	84	379.620	258
Measurement weights	70	394.245	272	14	14.625	0.404
Structural weights	64	399.563	278	6	5.318	0.504

Note: NPAR: number of parameters.

. First, considering that the measurement weights were equal, there was no significant difference in measurement weights (χ²_{14, 0.95%} = 14.625 < 23.685); then, proceeding to the next consideration, there was still no significant difference in the structural weight (χ²_6,95%= 5.318 < 12.592). In terms of results, the influence between the constructs did not differ by gender.

5. Discussion

5.1. Influence of Teaching for Contextual Competences for Sustainability

The result of this paper shows that teaching has a significant positive and direct influence on contextual competences for sustainability, consistent with research hypothesis 1 above. In other words, teaching methods promote the growth of students’ contextual competences for sustainability, which is consistent with the existing research results [59]. The results of the empirical study show that teachers choose appropriate teaching methods according to the teaching content and its characteristics, such as flexible case teaching, project-based teaching and problem-based teaching, which can lead students to actively think and effectively participate in the process of classroom teaching, thus contributing to the cultivation of students’ sustainable contextual competence. As the implementers of teaching, the use of teaching methods determines the teaching effectiveness. The classroom should be student-centered, encouraging students to explore, improving students’ willingness to participate in learning. Through project-based and question-based teaching methods, boring learning content is closely linked with practical engineering problems. Students increase knowledge while realizing the complexity of realistic activities. In group problem-solving, students integrate the knowledge involved and establish a comprehensive understanding of the problem based on listening to others. Furthermore, the classroom embraces experiential learning through hands-on activities and immersive simulations to deepen conceptual retention.

Teaching from a sustainable perspective emphasizes student-centered learning [3]. Therefore, teachers should adapt their teaching methods accordingly to emphasize students’ positive engagement. By fully utilizing project-based and inquiry-based approaches, teachers can integrate professional knowledge with real-world problems, allowing students to actively participate in discussions. Additionally, teachers should incorporate more authentic cases and practical experiences into the classroom to make knowledge relevant to students’ lives and enable them to experience the complexities of real environments.

5.2. The Mediating Roles of Student Interaction and Extracurricular Engagement

The results show that teaching not only directly affects contextual competences for sustainability but also indirectly affects contextual competences for sustainability through student interaction and extracurricular engagement, validating the validity of research hypothesis 2 above. In other words, extracurricular engagement and student interaction both play mediating roles in the influence of teaching on contextual competences for sustainability. This is consistent with existing research results [39]. Teachers adopting interactive teaching methods can provide more opportunities for student interaction. Collaboration and communication among peers not only helps students grow in their knowledge and skills, but also promotes the formation of students’ emotions and values through the “peer effect”. As students in the same class, there are many common topics among them. It is easier for students to understand and accept suggestions shared by a student in the same class, thereby changing their behaviors. The school, through a variety of teaching methods, enables students to have positive emotions about learning content, improves students’ interest in learning, and promotes students to actively participate in various extracurricular activities to improve their competences. Extracurricular activities involve various forms, such as competitions, project research, and volunteering. Through participating in real-life activities, students realize the diversity of the real environment. This improves their ability to think about their surroundings.

For the future, a combination of students’ subjectivity and teachers’ leadership is essential. Teachers should adopt interactive teaching methods to facilitate space for students’ effective communication. Furthermore, teachers should evaluate the outcomes of student discussions and provide encouragement, thereby stimulating their initiative and enthusiasm for further exploration. Throughout this process, teachers play a pivotal role in guiding course progression while also facilitating the interaction process by imparting experiences and skills to enhance students’ communication abilities. Engagement in extracurricular activities plays a pivotal role in students’ acquisition of competence. In order to enhance students’ interest and foster their active participation, teachers employ student-centered teaching methodologies. By adopting case teaching and project-based learning, courses are integrated with practical applications that enable students to identify real-world challenges encountered by enterprises. Moreover, teachers provide guidance to help students comprehend the significance of extracurricular engagement, thereby enabling them to select and participate in targeted activities.

5.3. The Chain Mediating Roles of Extracurricular Engagement and Student Interaction

The result of this paper shows that student interaction and extracurricular engagement have chain mediating roles in the influence of teaching on contextual competences for sustainability, which proves the positive influence of student interaction on extracurricular engagement. Specifically, as stated in research hypothesis 3, teaching directly affect the contextual competences for sustainability of undergraduates while also contributing to their development through a chain of mediating roles in student interactions and extracurricular engagement. This is consistent with existing research results [39]. Teaching can provide opportunities for interaction among students. If students engage in frequent collaborative problem solving, they can develop a profound comprehension of the knowledge they have acquired, thereby fostering their extracurricular exploration. Given that students are influenced by their peers during the process of socialization due to shared age and ideology, particularly through interactions with outstanding peers, less accomplished students gradually gain clarity regarding their areas of focus and endeavor.

In the classroom, teachers design teaching activities that facilitate student interaction, such as group discussions and collaborative assignments. This creates an interactive learning environment and fosters students’ trust and reliance. Additionally, teachers should actively understand the cultural characteristics of college students and the specific traits of their class, guiding students to effectively utilize peer support while cultivating a conducive atmosphere for student development.

6. Conclusions

Based on a review of the existing literature, a model of influencing factors of contextual competences for sustainability was constructed. A total of 786 engineering undergraduates were randomly surveyed to conduct a questionnaire survey to explore the effects of teaching, student interaction and extracurricular engagement on the contextual competences for sustainability. This paper found that teaching has a significant positive and direct influence on contextual competences for sustainability of engineering students, extracurricular engagement and student interaction mediate the influence of teaching on contextual competences for sustainability, and that student interaction and extracurricular engagement have chain mediating roles in the influence of teaching on contextual competences for sustainability. Based on the path test of the factors affecting the contextual competences for sustainability, this paper puts forward some training suggestions, which will help higher education institutions readjust their education plans. However, there some deficiencies in this study.

First of all, considering the influence of training years on the sustainable contextual competence of engineering undergraduates, for the purpose of controlling variables, the research object of this paper only involved senior undergraduates majoring in engineering. However, the cultivation of sustainable situational competence is a long-term process that cannot be focused only on talent cultivation results. The current study lacks attention to students at other levels, and this should be supplemented in subsequent studies in order to strengthen the scientific validity and reliability of the research conclusions.

Second, in terms of research variables, the perspectives of interaction variables and participation variables covered in this paper are not comprehensive enough. For example, teacher–student interaction and classroom participation were not involved. Teaching is a complex interpersonal process involving multiple subjects and multiple factors, and it is true that as researchers we cannot exhaust all the variables, but we can expand the perspective and comprehensiveness of a study within the limits of our ability. In subsequent studies, the variables measured can be added to facilitate the construction of a more complete model of factors influencing sustainable situational competence.

Finally, the sampling units of this study can be further expanded. Currently, the study samples are all from engineering college students in a certain region with roughly the same educational policies, hardware and software resources, economic development level and cultural background. However, these are all factors that may affect the development of students’ competence development, and the sampling range can be expanded in a subsequent study to investigate whether the development paths of engineering undergraduates’ sustainable situational competences are different in different regions or backgrounds.